2024 is the year of reckoning for AI

The AI ​​marketing hype, possibly started by OpenAI’s ChatGPT, has reached a fever pitch: investors and executives have stratospheric expectations for the technology. But the higher the expectations, the easier it is to disappoint. The stage is set to mark 2024 as the year for AI, as business leaders consider what AI can do In fact do it now.

The Gartner Hype Cycle shows that the peak of inflated expectations is followed by a period of disillusionment. Take the latest earnings from Alphabet, the parent company of Microsoft and Google. Alphabet reported a strong quarter: revenue of $86.2 billion, an increase of more than 13 percent from a year earlier. Its shares fell more than 7 percent the next day in reaction. Microsoft also gave similar information Its net income increased by 33 percent – and its shares also declined, though only by 2 percent.

Investors did not expect anything good. he was Hope Excellent, The tech sector is growing rapidly, with the so-called Magnificent Seven – Nvidia, Meta, Amazon, Tesla, Apple, Microsoft and Alphabet – leading the way. was well ahead of the market last year, Investments in AI have not yielded as good returns as the market had expected.

The business community’s expectations are a natural consequence of the breathless AI hype of the past two years Wall Street analysts are optimistic That AI is nothing less than “a generational change” that is “analogous to the advent of electricity.”

“We got frothy too early,” says Bernstein analyst Mark Shmulik. Some investors and analysts have analogized AI to the early Internet, which puts too much of the burden on the shoulders of AI companies. “But we have already gone through some of these cycles with crypto and the metaverse, which we quickly abandoned.”

There is no doubt that there is an appetite in the business world for AI tools. “Interest is universal across almost every industry,” says PwC’s Brett Greenstein. “However, it’s still early stages of adoption for everyone.”

Our business leaders expect returns on their AI investments in 2024, with 61 percent of CEOs surveyed expecting generative AI to boost the quality of their products and services. According to PwC’s annual survey, The PwC report warns of a “moment of truth” for trust in AI, where mistakes could lead to huge explosions; This quarter’s earnings show that the failure to deliver on promised productivity gains may be the most irritant.

Close observers like Greenstein believe IT may be the first to be widely adopted – particularly around software development, customer service and back-office functions such as processing inbound forms. This makes Microsoft’s co-pilot, the “AI companion”, an indicator for the overall field. A softer measure of Microsoft’s dominance is that, as “Googling” has increasingly become synonymous with web search, we’re talking about Copilot, not the offerings from other giants like Google, Salesforce or Oracle. says co-founder and general Adam Niewinski. Partner at OTB Ventures.

“Microsoft is ahead of Google and basically anyone else,” says Nivinsky. He told me that their partnership with OpenAI – about which, more in a minute – provides them with a source of growth in ways that may be more significant than just simple investment returns. “Everyone knows ChatGPT and OpenAI and everyone knows Copilot. But it took me a few seconds to remind myself that Google has Bard, because no one is actually using that name. (Bard’s name has since been changed to Gemini, as nothing more reliable could be said The astrological sign is most likely “Take you out of the complex of your own life.”)

Microsoft’s leadership sets it up for a positive feedback loop, where more users means more training, which means better data, which means more users. So seeing how Copilot is adopted can be a proxy for how AI is being implemented, says PitchBook’s Brendan Burke. “It’s an easy-to-use application that developers are testing to see if it can meaningfully help them, or is just a distraction,” says Burke.

Is Copilot really worth its price remains an open question, Certainly, AI didn’t help Microsoft with search – despite all the hype about incorporating AI into its Bing search engine. Its market share increased by less than half a percent, It looks like Microsoft is also trying to put Copilot to good use; The brand started on GitHub and moved into selling products and then Microsoft Office apps. even Microsoft’s Super Bowl ad is vague — If there is a lethal use case, the company hasn’t found it yet.

AI is expensive. Take OpenAI, for example; In December 2023, Its annual run rate was $2 billion, Because this is a figure that takes the previous month’s revenue and then multiplies it by 12, we know that this means OpenAI made approximately $167 million that month. Yet it is loss-making and will likely need to raise “tens of billions of dollars” to continue. financial Times informed of. Sam Altman is the CEO of OpenAI. is wanting billions dollars in investment To completely reshape the chip industry. During this, Chatgpt development has stopped,

During the era of zero interest rates, big tech can plow endless money into their favorite projects – CEO Mark Zuckerberg’s little adventure in the metaverse At least $46.5 billion lost since 2019, Luck Reported last October. Perhaps if we were still in that era, a company like Google could invest money in AI. “I don’t think Google can afford to spend lavishly on these initiatives,” says Shmulik. “We are living through a period where investors increasingly care about profitability.” This doesn’t seem like very good news to me for Geminis, which so far seems to be A slightly better version of Google Assistant and is reduced by ChatGPT, According to tom’s guide,

Even Google is cutting costs (12,000 jobs last year and more in 2024), and the bar for funding is likely to be higher in 2024. The challenge for AI startups now is to build sustainable business models as well as bring AI into areas that have not yet been disrupted, Burke told me. And the high valuations relative to revenues assigned to these companies show that VCs expect them to become tech giants in the long term.

According to a note from PitchBook’s Burke, VCs have tightened their purse strings and even the AI ​​field has been affected. Excluding deals made by Microsoft and Amazon, which are not quite traditional VC investments, only $7.9 billion had been invested by VCs in the sector as of last November – meaning the same level of spending as in 2021.

There may also be fewer exits for VCs investing in the AI ​​sector. As rumblings about tech antitrust have grown louder, concerns have grown about whether acquisitions might pan out — Adobe’s attempt to buy Figma, after all, didn’t pan out. “Many people already believe that these megacap companies have too much power, and AI is clearly a sensitive point right now,” says Angelo Zino, technology analyst at CFRA Research. “That’s why you haven’t seen a complete takeover.”

Instead, AI fundraising is set to surge in 2023 as large-cap companies pour money into the sector. “We won’t see a deal of the scale between Microsoft and OpenAI or Amazon and Anthropic this coming year, just as those companies have established their leadership in this area,” Burke says.

However, AI companies have a compelling pitch for anyone interested in using their services. For example, ServiceNow CEO Bill McDermott has said that It can make engineers 50 percent more productive, Scott Kessler, head of technology at Third Bridge, says this is a tremendous cost savings. Theoretically, the demand for AI should be very high.

But real adoption and real benefits have yet to materialize. As yet, It’s not clear that Microsoft’s CoPilot is useful enough to justify its cost.According to wall street journal, The uptake has been slower than expected; Microsoft had to drop its initial requirement that companies sign up for at least 300 subscriptions. It also appears that usage is declining a month after the initial surge. But don’t worry, Microsoft has a plan to address this: It’s going to create more alerts.

PwC’s Greenstein says that right now, most AI companies are making money through premium pricing for better services. But that could potentially evolve into outcomes-based pricing, he says. “The idea of ​​paying on results – you know, ‘I saved you money, I get something, you get something, everyone wins’ – it’s a really interesting model in AI, because there are so many things that [where] You can be assured of results,” he says. “And I think these will be very compelling commercialization models.”

This can help avoid disappointment. Because if AI tools still require human review, they can’t be any faster or better than if a person did it in the first place. Greenstein spoke to me on the phone excited about using AI to summarize information, but he and I had very different experiences. Perhaps the AI ​​does a good job of summarizing forms, assuming it doesn’t make any errors.

But I’ve tried using AI to summarize podcasts I don’t particularly want to listen to, then listened to them to compare and see how well the AI ​​works. I think that although the AI ​​summaries are technically correct, neither Gemini nor ChatGPT have enough context (or social skills?) ) Is. The background knowledge required for anything was beyond the AI’s capabilities. Using AI actually increased the time it took me to do my work – the time it took to ask for an AI summary and read it.

There is even OpenAI trying to push back on the hype, In December, OpenAI chief operating officer Brad Lightcap told CNBC that he had to explain to people that AI can’t dramatically cut costs or bring back growth for struggling companies. Morgan Stanley’s AI chatbot is being ignored by money managers because people want to talk to other people, Information informed of. News operations that have attempted to replace journalists with AI-written articles have faced backlash in the form of articles gone wrong, unpleasantOr Useless,

So it looks like 2024 is where the rubber will hit the road. If there are real use cases for large language models that save businesses money, then perhaps AI will be on its way to sustainability. But if these devices come into widespread use and move bad publicity, lawsuitsAnd in congressional hearings, with minimal productivity gains, a trough of disillusionment could set in – and it could be very deep indeed.